I think the answer is no, but I was curious to know if Vision Framework can currently perform Connected-component labeling (AKA blob extraction, region labeling, blob discovery, or region extraction)?
https://en.wikipedia.org/wiki/Connected-component_labeling
Being able to detect a region within a binary, greyscale, or even color image would be super useful – and it's definitely a good candidate for Vision because it's in the realm of computer vision.
However, I'm not sure if Vision Framework supports it. Perhaps we'll see it as a future enhancement to the framework if it doesn't do it now?